Structure-Activity Relationship Studies at the Benzodiazepine Receptor

Benzodiazepine receptor (BZR) ligands exert a contin- uous profile of pharmacological activities,lP2 ranging from. (i) agonists, which enhance y-amino...
1 downloads 0 Views 622KB Size
J. Med. Chem. 1993,36, 1669-1673

1669

Notes Structure-Activity Relationship Studies at the Benzodiazepine Receptor (BZR): A Comparison of the Substituent Effects of Pyrazoloquinolinone Analogs? R. Ian Fryer,*l*Puwen Zhang? Roberto Rios,*Zi-Qiang Gu,O Anthony S. Basile,s and Phil Skolnickt Department of Chemistry, Rutgers, The State University of New Jersey, Newark, New Jersey 07102, and Laboratory of Neuroscience, National Institute of Diabetes and Digestive and Kidney Diseases, NIH, Bethesda, Maryland 20892 Received January 14, 1993

The synthesis of a series of 2-phenylpyrazolo[4,3-clquinolin-3-one derivatives and their in vitro biological evaluation as ligands for the benzodiazepine receptor are described. The in vitro activities, as determined by an analysis of GABA shift ratios, and binding affinities of these compounds to BZRare compared in terms of the electronic, lipophilic, and steric effect changes of their substituents. Benzodiazepine receptor (BZR) ligands exert a continuous profile of pharmacological activities,lP2ranging from (i) agonists,which enhance y-aminobutyric acid (GABAA) gated C1- current to induce anticonvulsant, hypnotic, muscle relaxant, and anxiolyticeffects; (ii) inverseagonists, which are proconvulsants, convulsants, and anxiogenics; and (iii) antagonists, which bind to the BZR but have no pharmacological effects. The detailed tertiary structure of the BZ/GABA receptor supramolecularcomplex remains obscure even though recent reports suggested that at least three distinct cDNA’s must be coexpressed in vitro to obtain a fully functional chloride channeL3 Our interest in pyrazolo[4,3-c]quinolin-3sne compounds, such 88 CGS-9896, CGS-9895, and CGS-8216 (4, 7,15 respectively, Table I), resulted from the report that these compounds bind to BZR with the very high affinity and their different pharmacological profiles are related only to the nature of the substituent in the para position of the 2-phenyl ring.4 The difference in the pharmacological activities for the pyrazolo[4,3-clquinolin-3-one compounds has been separately discussed in terms of either lipophilic and electronic6or steric effects6of the substituents. Since the marked difference of these biological profiles could be attributed to any factor or to a combination of them, it does not seem reasonable to discuss some effects without concomitant consideration of the others. In an attempt to systematically evaluate these substituent effects on receptor affinity and on the activity profile of pyrazolo[4,3-clquinolin-3-ones, a series of pyrazoloquinolinone compounds was synthesized. These compounds were designed on the basis that the relative lipophilicity of both the fused benzene ring and the 2-phenylring can be altered by an appropriate substitution. On the other hand, a change of the substituent position on the same aryl ring would not be expected to affect lipophilicity’ in any substantial manner, although such changes would be expected to alter both the electronic properties and the spatial orientation of the substituents in the molecules. t Presented in part at 201st National Meeting of American Chemical Society, Atlanta, GA,April 1991; Med. Chem. No. 128. Rutgers, The State University of New Jersey. 8 NIH.

Table I. Pyrazoloquinolinonesa

H

no.

R

R‘

no.

R

4 5 6 7 8 9

H H H

p-C1 m-C1 0-c1 p-OMe m-OMe o-OMe

10 11 12 13 14 15

&C1 8-C1 841 &OMe 7-OMe

R’ H pC1 0x1

H H H H H H H a In three instances (during the preparation of compounde 6.12, and 14), the initially formed hydrazines 16 were isolated prior to cyclization of the pyrazole ring. These intermediates were readily characterized by the ‘H-NMFt spectroscopy. No attempt waa made to further purify these hydrazines, and they were cyclized to their respective pyrozoloquinolinones by heating in Dowtherm A. ~

~~

R-&”a COEl

16

Chemistry

All pyrazolo[4,bcIquinolin-8onecompounds were made using slightly modified literature procedure^.^^^ In brief, the substituted anilines were refluxed with diethyl (ethoxymethy1ene)malonate in Dowtherm A to give 4-hydroxy-3-quinolinecarboxylicacid esters 2 via thermal ring closure. Treatment of 2 with phosphorus oxychlorideled to the corresponding chloro derivatives 3, which were condensed with the appropriately substituted phenylhydrazines in xylene or Dowtherm A to afford 4-15 (Scheme I and Table I). The known intermediate compounde 2a-d and 3a-d had melting points consistent with literature values, and their spectra were compatiblewith the assigned structures.8 The find compounds were purified by recrystallization from ethanol. The three known compounds (4,7, and 15) were also resynthesized as biological standards and had melting points and exhibited spectral properties consistent with literature value^.^

0022-2623/93/1836-1669$04.00/00 1993 American Chemical Society

Notes

1670 Journal of Medicinal Chemistry, 1993, Vol. 36, No. 11

Scheme I. Synthesis of Pyrazoloquinolinone Analogues" OH

l(a-d)

CI

2(a-d)

3(a-d)

Figure 1.

wo N-N

Xa-d)

"'- Rh

I

a: 1 (R=H a4 b: R=p-CI c: R = ~ . o M ~ d , R=m.OMe

2, 3 (a-d) a: R=H b: R=6-CI c: R=6-OMe

d. R=l-OMe

li 4-15

Reagents and conditions: (i) diethyl(ethoxymethylene)malonate, Dowtherm A; (ii) Poc13; (iii) phenyls hydrazine, ( 0 ; m-,or p-chloroand -methoxyphenyl)hydrazine,xylene, or Dowtherm A.

Table 11. Potencies and GABA Shift Ratio of Pyrazoloquinolinone Analogs Compared with Classical 1,4-Benzodiazepines" ICs0 (nM)b)

I

GABA shift ratio

compound 4(CGS-9896) 5 6 7(CGS-9895) 8 9 10 11 12 13 14 15(CGS-8216)

0.56 i 0.04 (0.62) 3.90 f 0.31 70.0 i 2.3 0.38 f 0.03 (0.06) 0.58 i 0.07 990 f 46 0.24 f 0.05 4.00 i 0.15 8.00 i 0.23 0.38 i 0.03 2.10 f 0.3 0.13 i 0.01 (0.15)

0.94 (0.86) 1.1 NDe 1.1 1.47 1.34 0.91 0.97 0.99 (0.94)

diazepam flumazenil

20.6 i 2.3 (15.1) 1.82 i 0.28 (2.3)

1.78 (1.82) 1.00 (1.00)

1.07 (1.32) 1.7

activities partial agonistc

1.08

partial agonist/ antagonistc

partial inverse agonistd agonist antagonist

The data in the parentheses were KDvalues taken from ref 5, and GABA shift ratio in the parentheses were taken from the same reference, I C s values reported here were determined at a 1 nmol final concentration of radioactive ligand. Resulta reported as 95% fiducial limits. See ref 17. See ref 18. e GABA ratio was not determined because of ita low potency.

*

Results and Discussions Studies have indicated that the GABA shift ratio (the ratio of IC50 without GABA/IC60 with GABA) can be related to9 and has been extensively employed as a predictor oflo the biological spectrum of BZR ligands, in which GABA modulates the binding of BZR ligands, enhancing that for agonists (ratio > 1.01, not affecting that for antagonists (ratio = 1.0) and diminishing that for inverse agonists (ratio C 1.0)." Using the GABA shift ratio of the standard BZR ligands (diazepam and flumazenil) shown in Table I1 as a basis for comparison, compounds 5 and 6, isomers of CGS-9896 (4), have the same relative lipophilicity as that of 4 and show the character of an agonist-like and a partial agonist/antagonist-like pharmacological profile, respectively. Compound 6, which has very similar lipophilic and electronic properties as compound 4, seems to retain the pharmacological profiles of the parent compound. The reduction of binding affinity of 6 is attributed to the steric effects caused by the different substituent position. Compound 5, on the other hand, possesses different electronic effects and different steric effects from 4, thus altering both its potency and its pharmacological profile. Since the steric effect (compound 6) reduces only the binding affinity, the profile change of 5 from partial agonist/antagonist to the

character of agonist may be ascribed, at least partially, to a difference in the electronic effect of the parasubstituent from the metasubstituent. The same type of combined electronic and steric effect on the potencies and pharmacological profiies is also observed in the methoxy isomers (7, 8, and 9) which are related to CGS-9895 (7). These results imply that the potencies of these ligands are modulated by the substituent position on the 2-phenyl ring, while their pharmacological profiles might be related to the electronic effects of the substituents. Thus, the findings presented here do not agree with those described by Shindo and co-workers,6in which the pharmacological activities of substituted 2-thienylpyrazoloquinolinones, the analogs of 2-phenylpyrazoloquinolinones,were predicted based only on the relative distance of the substituents on a thiophene ring measured from the axis along the N-C bond connecting the pyrazolo and thiophene rings. It should be noted here that as the substituents on the 2-phenyl ring are moved from the para position through to the ortho position, steric hindrance can possibly induce rotation about the bond between N2 and the 2-phenyl ring, causing an increasing out of plane twist of the 2-phenyl ring. The lack of binding affinity of compound 9 is believed to be such an example caused by a twist of the 2'methoxyphenyl ring. A second series of compounds (10, 13, and 14) was synthesized in which the B ring was kept unsubstituted, while the A ring was substituted either by Cl or by OCHa. Substitution at the 8-position was considered to be comparable to the para position of the 2-phenyl ring if the orientation of the molecule is inverted within the receptor pocket (Figure 1). It was reasoned that a pyrazoloquinolinone with the same relative lipophilicity between the A and B rings as that for the partial agonist/antagonist 7 could be synthesized by the removal of the methoxy (T value around _0.0212)substituent from the 2-phenyl group of the compound 7 and adding a chlorine atom (?r value around 0.7012)to the 8-position of the parent compound 15, i.e. 10.13 Thus, 10 would have the opposite relative lipophilicity between the A and B rings compared with 4 and the same relative lipophilicity as for compound 7. Both potencies and pharmacological profiles of 4 and 10, as shown by GABA ratios, appear to be the same. This suggests that the binding orientations of the substituted rings within the receptor sites must coincide. In other words, the results seem to infer that one of these two molecules would bind in an inverted configuration within the receptor pocket (Figure 1)in order for the substituents to be superimposed. Similarly, compounds 7 and 13 (methoxy analogs of 4 and 10) also exhibit the same potencies and the same pharmacological profiies despite the opposite relative lipophilicity between their A and B rings. These findings would seem to imply that ligand orientation within the receptor active sites can be modulated by the relative lipophilicity of the aromatic rings in pyrazoloquinolinones. The dichloro-substituted compounds 11 and 12 have the "GABA-positive"character indicating an agonist-like

Notes

Journal of Medicinal Chemistry, 1993, Vol. 36, No. 11 1671

profile and have the same potencies at the BZR. The pharmacological profiles and potencies of 11and 12 appear to be different from those of their A-ring unsubstituted analogs (4 and 6). The latter show a partial agonist/ antagonist profile, and their potencies exhibit a difference of 2 orders of magnitude. Aa mentioned above, moving the C1 on the freely rotating phenyl ring from 4'- to 2'position resulted in a remarkable loss of binding affiity, presumably caused by an increase in out of plane rotation. This might well indicate that this aromatic ring is acting as the molecular determinant which has the ?r-raromatic stacking interactions with the BZR for the monosubstituted analogs. For the corresponding disubstituted analogs (11 and 12) this loss of affinity was not apparent for the 2'-Cl-substituted analog. This result can be rationalized as being due to the fact that it is now the fused benzene ring that acts as the molecular determinant. Thus, the findings presented here are at variance with those described by Villar et al.5 in which it is proposed that the substituted 2-phenyl ring of pyrazoloquinolinonesacts as the molecular determinant, interacting with BZR to d e f i e an agonist profile. The enol configuration of the vinylogous amide was required by Villar's model in order to explain the agonist activity 0fCGS-9896 (twoimine nitrogen atoms are defined as binding sites). From the NMR (in DMs0-d~)and FTIR (KBr) data for the pyrazoluquinolinones described in this paper, no evidence for the existence of enol formation was ever detected. This is consistent with reported data with a series of 2-thienylpyrazoloquinolinones6and 2-phenylpyrazoloisoquinolinol~.~~ Thus, the physical data for these agonist ligands indicate only the carbonyl form of the amide, which again is at variance with Villar's required model for BZR agonists.

were performed at Hoffmann-La Roche and were all within f0.4% of calculated values for carbon, hydrogen, and nitrogen. All chemicals were purchased from the Aldrich Chemical Co. RadioligandBinding. The affinity of the compounds for the central benzodiazepine receptor was assessed by using a modification of the previously described techniques.ls Sprague-Dawley rats were decapitated, and the brains were rapidly removed and placed in 320 m o l of sucrose (0-4 "C) before dissection. The cortex was homogenized in 50 volumes of 50 mM Tris-citrate buffer (pH 7.4) using a Polytron (Brinkmann Instruments) at a setting of 6.5 for 15 s. The homogenate was centrifuged at 2oooOg (0-4 "C) for 20 min, and the pellet was resuspended in an equal volume of Tris-citrate buffer. This "washing" procedure was repeated five times. After the last wash, the pellet was resuspended in 20 volumes of buffer and stored at -80 OC for no more than 30 days before use. Before assay, the tissue preparation was thawed, and a 50-pL aliquot (containing about 0.12 mg of protein) was added to each assay mixture, which also contained 50 pL of r3H1Ro15-1788(final concentration, 1nM) and 50 pL of Ro14-7437for nonspecific binding (finalconcentration, 10 pM). Sufficient Tris-citrate buffer was used to bring the final assay volume to 0.5 mL. The assay was terminated after incubation (1 h at 0-4 OC) by rapid fitration over Whatman GF/B fiiter strips using a Brandel M-24R filtering manifold. Samples were washed twice with 5 mL of cold buffer. The filters were then placed in vials and 4 mL of scintillation fluid was added. In order to determine the GABA shift ratio, a similar procedure was carried out in the presence of NaCl (400 mM) and GABA (200 mM). The radioactivity retained by the filters was measured in a Beckman LS 5802 liquid scintillation spectrometer. The concentration of test compounds which inhibit the specificbinding of [3HlRo15-1788to BZR in rat cortex by 50% (ICw, nM) were derived from displacement curves with 10 concentrations of each compound, each assayed in triplicate. General Procedures for the Preparation of Pyrazoloquinolinones. A mixture of compounds 3a-d and the appropriate analog of phenylhydrazine (approximately 1.2 molar equiv) in xylene or Dowtherm A was heated at the boiling point temperature of xylene or in those cases requiring higher temperature (at about 180 O C ) in Dowtherm A. After 1h, the reaction mixture was cooled to room temperature and filtered. The yellowish solid was washed several times with hexane and then several times with acetone. The solid was then dissolved in the hot ethanol and filtered to remove undissolved impurities, and the compound was crystallized by cooling. In cases when the compound had poor solubility in ethanol, a few drops of 3 N sodium hydroxide solution was added to the ethanol solution, which was then filtered, and the filtrate was carefully neutralized with 1N HC1 (to avoid precipitation) to pH 7 (pH paper). The solution was cooled and the product allowed to crystallize. Filtration afforded the crystalline product which was washed well with distilled water and dried in vacuo. Preparationof 2a-d and 3a-d. Compounds 2a-d and 3a-dwere prepared according to the literature procedure.8 The melting points of these compounds were consistent with the literature values.6

Conclusion The present studies demonstrate that the affinity of pyrazoloquinolinoneligands to BZR can be correlatedwith the steric effects due to substituents on the 2-phenyl ring in these compounds. Furthermore, their pharmacological profiles seem to be modulated by the electronic effects of these substituents. The relative lipophilicity appears to define the aromatic ring determinant of pyrazoloquinoh o n e s binding to the BZR. Nonetheless, the correlation between the relative lipophilicity and the aromatic ring determinant is not discernible as described by Villar et ales It is also apparent that the rationalization of the intrinsic activity of these compounds based solely on steric effects6of the substituents on the 2-phenyl ring is not a general phenomenon. The consideration of any one substituent effect (Le., lipophilic, electronic, or steric) without concomitant consideration of the others is inadequate as a predictor of biological activity for the pyrazoloquinolinone ligands. Further work is needed in order to bettar d e f i e a predictive model for the pharmacophores of pyrazoloquinolinone compounds.

Experimental Section Melting points were determined with the Mel-Temp apparatus and are uncorrected. Infrared spectra were obtained by usingthe Mattaon Polaris FT-IR spectrometer with KBr pellets. lH NMR spectra were taken either on a Bruker wP-200SY or on a Varian 400 Fourier transform spectrometers using DMSO-ds solvent. Microanalyses

Notes

1672 Journal of Medicinal Chemistry, 1993, Vol. 36, No. 11

2-(3-Chlorophenyl)pyrazolo[4,3-c]quinolin-3-one (5). A mixture of 3-carbethoxy-4-chloroquinoline(3a)(2.0 g, 9 mmol) and (3-chloropheny1)hydrazine(2.0 g, 13 mmol) in 60 mL of xylene was heated according to the general procedure. Recrystallization of product from the solvent system as described in the general procedure afforded 5 (1.62 g, 61% ) as yellow crystals: mp 325-326 "C (lit.16mp 336-337 "C); lH NMR (DMSO-&) 6 12.9 (br, NH, D20 exchangeable), 8.8 (d, 2 H, J = 6.5 Hz), 8.35 (s, 1 H), 8.2 (t, 2 H, J = 6.7 Hz), 7.7 (m, 2 H), 7.55 (dd, 1H, J = 2.8, 8.0 Hz), 7.45 (t,1H, J = 8.2 Hz); IR (KBr) 3377.1,3135.1, 2975.9, 1630.7, 1588.2, 769.5 cm-l. Anal. (Cl6HlOClN30.0.5H20) C, H, N. 2-(2-Chlorophenyl)pyrazolo[4,3-c]quinolin-3-one (6). Treatment of 3a (2.0 g, 9 mmol) with (2-chloropheny1)hydrazine (2.0 g, 13 mmol) in 50 mL of xylene solution under reflux for 1.5 h afforded the corresponding uncyclized intermediate 16 formed by nucleophilic attack of hydrazine at the 4-position of quinoline. This structure was confirmed by IH NMR spectroscopy. The intermediate was then reheated in 60 mL of Dowtherm A at about 180 "C for 30 min. Removal of solvent afforded 6 (1.48 g, 56%) as bright yellow crystals: mp 330-331 "C (lit.16 mp 336-339 "C); lH NMR (DMSO-&) 6 12.95 (br, NH, D2O exchangeable), 8.75 (d, 1 H, J = 6.4 Hz), 8.1 (dd, 1 H, J = 2.0, 7.6 Hz), 7.5-7.7 (m, 7 H): IR (KBr) 3441.7, 3060.8,2840.9,1615.2,1455.2,758.9cm-l. Anal. (C16H10ClN30) C, H, N. 2 4 3-Methoxyphenyl)pyrazolo[4,3-c]quinolin-3one (8). A solution of 3a (2.0 g, 9 mmol) and (3-methoxypheny1)hydrazine (2.0 g, 14 mmol) in 50 mL of xylene was heated for 1h and after workup afforded 1.49 g (57% ) of 8 as yellow crystals: mp 248-249 OC; lH NMR (DMSO&) 6 13.0 (br, NH, D20 exchangeable), 8.8 ( 8 , 1 HI, 8.25 (d, 1H, J = 7.6 Hz), 7.5-7.9 (m, 5 H), 7.6 (t, 1 H, J = 8.1 Hz), 6.8 (dd, 1H, J = 1.5,8.4 Hz), 3.85 (s,3 H); IR (KBr) 3487.0,3104.2,2957.3,2864.1,1650.9,1600.8,1453.2,1234.3, 759.9 cm-l. Anal. (Cl~H13N302)C, H, N. 2-(2-Methoxyphenyl)pyrazolo[4,3-c]quinolin-3one (9). A mixture of 3a (2.0 g, 9 mmol) and (2-methoxypheny1)hydrazine (2.0 g, 14 mmol) in 50 mL of xylene was heated under reflux as described above and after workup afforded 1.3 g (50%) of 9 as yellow needles: mp 305-307 "C; lH NMR (DMSO-&) 6 13.1 (br, NH, D2O exchangeable), 8.65 (8, 1H), 8.2 (d, 1H, J = 7.6 Hz), 7.25-7.75 (m, 5 H), 7.2 (d, 1H, J = 8.3 Hz), 7.12 (t, 1 H, J = 7.5 Hz), 3.75(s,3H);IR(KBr)3479.4,3401.2,3056.9,2839.0,1621.1, 1579.5,1454.2,763.76 cm-l. Anal. (C17H13N30~0.5H20) C, H, N. 8-Chloro-2-phenylpyrazolo[ 4,3-c]quinolin-3-one(10). A solution of 2.0 g (7 mmol) of 4-chloro-6-chloro-3carbethoxyquinoline (3b) and 1 mL (10 mmol) of phenylhydrazine in 60 mL of xylene was heated for about 30 min. After workup 10 was obtained as yellow crystals (1.07 g, 52%): mp >400 "C dec; 'H NMR (DMSO-dd 6 12.95 (br, NH, D2O exchangeable), 8.9 (s, 1 H),8.1 (m, 3 H), 7.85 (m, 2 H), 7.5 (t, 2 H, J = 8.8 Hz), 7.2 (t, 1 H, J = 7.3 Hz); IR (KBr) 3466.8,3184.3,3054.7,2939.3,2839.0, 1625.9,1450.3,758.9 cm-l. Anal. ( C I ~ H ~ O C ~C, NH, ~ ON.) 8-Chloro-2-(4-chlorophenyl)pyrazolo[ 4,3-c]quinolin-3-one(11). A mixture of 3b (2.69 g, 10 mmol) and (4-chloropheny1)hydrazine(1.7 g, 12 mmol) was heated in 60 mL of xylene for 30 min, and 11 precipitated as a yellow solid. Crystallization afforded 1.0 g (30%) of the product as yellow crystals: mp 366-369 "C; lH NMR (DMSO-&)

6 12.85 (br, NH, D2O exchangeable), 8.75 (d, 1H, J = 4.9 Hz),8.25 (d, 2 H, J = 9.0 Hz), 8.15 (d, 1 H, J = 1.9 Hz), 7.75 (m, 2 H), 7.45 (d, 2 H, J = 8.9 Hz); IR (KBr) 3421.5, 3095.6,2938.4,1629.7,1491.8,823.5cm-l. Anal. (CleHgCl2N30) C, H, N. 8-Chloro-2-(2-chlorophenyl)pyrazolo[ 4,3-c]quinolin-3-one(12). A mixture of 2.7 g (10 mmol) of 3b and 1.7 g (12 mmol) of (2-chloropheny1)hydrazinewas heated in 50 mL of xylene for 1h. Workup afforded the uncyclized intermediate (confirmed by 'H NMR spectrum) which was reheated in 60 mL of Dowtherm A at a temperature of about 260 "C for 1 h in order to effect cyclization. Workup affored 12 as yellow crystals (1.98 g, 60%): mp 379-383 "C; lH NMR (DMSO-&) 6 12.9 (br, NH, D2O exchangeable), 8.8 (s, 1H), 8.05 ( 8 , 1H), 7.75 (s,2 H), 7.65 (m, 1H), 7.45 (m, 3 H); IR (KBr) 3435.0, 3124.5, 2940.3, 1666.4,1607.6,1440.7,1271.0,759.9 cm-l. Anal. (C16HDC12N30) C, H, N. 8-Methoxy-2-phenylpyrazolo[ 4,3-c]quinolin-3-one (13). The reaction of 2.7.g (10 mmol) of (4-chloro-3carbethoxy-6-methoxyheny1)hydrazine (3c)with 2 mL (20 mmol) of phenylhydrazine in 50 mL of Dowtherm A at 250 "C for 30 min afforded 13. Recrystallization of 13 from ethanol afforded yellow crystals (1.4 g, 48%1: mp 322-325 "C; lH NMR ( D M s 0 - d ~6) 12.9 (br, NH, D2O exchangeable), 8.7 (d, 1H, J = 5.3 Hz), 8.27 (d, 2 H, J = 8.0 Hz), 7.7 (d, 1H, J = 9.1 Hz), 7.6 (8, 1H), 7.45 (t, 2 H, J = 7.8 Hz), 7.25 (d, 1H, J = 9.1 Hz), 7.2 (t, 1H, J = 7.1 Hz), 3.95 (s,3 H); IR (KBr) 3445.6,3064.7,2923.9,2849.6, 1632.6,1439.8,1239.2,757.8cm-'. Anal. (C17H13N302) C, H, N. 7-Methoxy-2-phenylpyrazolo[ 4,3-c]quinolin-3-one (14). A mixture of 2.0 g (7 mmol) of 4-chloro-7-methoxy3-carbethoxyquinoline (3d) and 1 mL (10 mmol) of phenylhydrazine was heated under reflux in 50 mL of xylene giving only the uncyclized intermediate. Compound 14 was obtained by reheating the intermediate in 60 mL of Dowtherm A at 240 "C for 30 min. Purification by crystallization afforded 1.6 g (76%) of 14 as yellow crystals: mp 297-299 OC; lH NMR (DMSO-&) 6 12.75 (br, NH, D2O exchangeable), 8.7 (d, 1H, J = 6.0 Hz), 8.25 (d, 2 H, J = 8.6 Hz), 8.15 (d, 1 H, J = 9.6 Hz), 7.45 (t, 2 H, J = 7.7 Hz), 7.2 (m, 3 H), 3.8 (s,3 H); IR (KBr) 3340.5, 3117.7,2963.4,1630.7,1469.7,1250.8,1032.8,865.9,766.6 cm-l. Anal. (C17H13N302) C, H, N. Acknowledgment. We wish to thankDr. N. W. Gilman for very useful discussion and Dr. F. Scheidl of HoffmannLa Roche Inc. for microanalytical data. The financial support from Hoffmann-La Roche Inc. and the Excellence fellowship for Rutgers University (P.Z.) and NIH-MBRS (5 SO6 GM08223-09, R.R.) are gratefully acknowledged. References (1) Haefely, W.; Kyburz,E.;Gerecke,M.;Mohler, H. Recent Advances

in the Molecular Pharmacology of BenzodiazepineReceptors and in the Structure-activity Relationships of Their Agonists and Antagonists. In Advances in Drug Research; Testa, B . , Ed.; Academic: Orlando, 1985; Vol. 14, pp 165-322. (2) Braestrup,C.; Honore, T.;Nielsen, M.; Petersen, E. N.;Jensen, H. Ligands for BenzodiazepineReceptorswith Positive and Negative Efficacy. Biochem. Pharmacol. 1984,33,859-862. (3) Pritchett, D. B.; Sontheimer, M.; Shivers, B. D.; Ymer, S.; Kettenmann, H.; Schofield, P. R.; Seeburg,P. H. Importance of a Novel GABAAReceptorSubunitfor BenzodiazepinePharmacology. Nature 1989,338, 582-585. (4) Yokoyama, N.; Ritter, B.; Neubert, A. D. 2-Arylpyrazolo[4,3-c]quinolin-3-ones: Novel Agonist, Partial Agonist and Antagonist of Benzodiazepines. J . Med. Chem. 1982, 25, 337-339.

Notes (5) Villar, H. 0.; Uyeno, E. T.; Toll,L.; Polgar, W.; Davies, M. F.; h e w , G. H.Molecular D e t e r m k t a of B e d i p i n e Receptor Affiiities and Anticonvulsant Activities. Mol. Pharmacol. 1989, 36,589-561. (6) (a)Takada, S.; Shindo, H.;Sasatani, T.; Mateushita, A.; Eigyo, M.; Kawasaki, K.; Murata, S. A New Thienylpyrazoloquinoline: a Potent and Orally Active Inverae Agonist to Benzodiazepine Receptors. J. Med. Chem. 1987, 30, 454-455. (b) Shindo, H.; Takada, S.; Murata, S.; Matsuehita, A.; Eigyo, M. Thienylpyrazoloquinolines with High Affiiity to Benzodiazapine Receptors: Continuous Shift from Inverse Agonist to Agonist Propertiee Depending on the Size of the Alkyl Substituent. J. Med. Chem. 1989,32, 1213-1217. (7) (a) Ghose, A. K.; Pritchett, A.; Crippen, G.M. Atomic Physicochemical Parameters for Three Dimensional Structure Directed Quantitative Structure-Activity Relationships I11 Modeling Hydrophobic Interactions. J. Comput. Chem. 1988, 9, 80-90. (b) Leo,A.; J h s c h , C.; Elkins, D. Partition Coefficiente and Their Uses. Chem. Rev. 1971,71,525-616. (8) Kaslow, E. E.; Clark, W. R. Quinolinemethanola. J. Org. Chem. 1953,18, 55-58. (9) Chan, C. Y.; Farb, D. H. Modulation of Neurotransmitter Action: Control of the y-aminobutyric Acid Fhponse Through L e Benzodiazepine Receptor. J. Neuroeci. 1985,42365-2373. (10)(a) Sanger, D. J. GABA and the Behavioral Effecte of Anxiolytic Druga. Life Sci. 1985,36,1503-1513. (b) Skerritt, J.H.; Johnaton, G. A. R.; Braeatrup, C. Modulation of CABA Binding to Rat Brain Membranes by Alkyl fl-carboline 3-carboxylate Esters. Eur. J. Pharmacol. 1983,86,294-301. (11) Braestrup, C.; Nielsen, M.; Honore, T.; Jensen, L.; Petersen, E. Benzodiazepine Receptor Ligands *th Positive and Negative Efficacy. Neuropharmacology 1983,22,1451-1457.

Journal of Medicinal Chemistry, 1993, Vol. 36, No. 11

1673

(12) Craig, P. N. Guidelines for Drug and Analog Design. In The Back of the Medicinal Chemistry, 4th ad.; Wolff, M. E., Ed.; Wiley-interscience: New York, 1980; pp 331-348. (13) Villar and coworkers assigned the 2-phenyl ring as the most lipophilic aromatic ring for compound 15. For the methoxy (1 value -0.02) substituted compound (7), the fused benzene ring was defiied as the moat lipophilic aromatic ring. This changes in the relative lipophilicity seemed to be attributed to net decrease of r value of 0.02 due to methoxy substitution on 2-phenyl ring. Therefore, we expected that by removal of the methoxy substituent from the 2-phenyl group plus addition of a chlorine atom (1value around 0.7) to the &position of the parent compound 15 would eervetomake the fused benzeneringofcompound 1Omorelipophilic than that of compound 7. (14) Michael, S. A.; Skolnick, P.; Cook, J. M. Synthesis of Novel 2-phenyl-2H-pyrazolo[4,3-c]isoquinolin-3-01~: Topological Comparisons with Analogues of 2-Phenyl-2,5-dihydropyrazolo[4,3-c]quinolin-3(3H)-onesat BenzodiazepineReceptors. J.Med. Chem. 35,368-374. (15) (a) Squires, R.; Braestrup, C. Benzodiazepine Receptors in Rat Brain Nature 1977.266. 732-734. (b) Molher. H.:Richards. J. Agonist and Antagonist Benzodiazepine Receptor Interaction In Vitro. Nature 1981,294,763-765. (16) Yokoyama, N. U.S. Patent4 312 870,1982; Chem. Abetr. 1982,96, 181283. (17) Bennett, D. A. Pharmacology of the Pyrazolo Type Compounde: Agonist, Antagonist and Inverse Agonist Actions. Physiol. Behau. 1987,41, 241-245. (18) Jenaen, L. H.;Petersen, E. N.; Braeetrup, C. Audiogenic Seizures in DBA/2 Mice discriminate Sensitively Between Low Efficacy BenzodiazepineReceptor Agonists and Inverse Agonista. Life Sci. 1983,33,393-399.

mz,